Emerging and re-emerging infections are widely acknowledged to be an urgent and growing global threat to human health. Nowhere is this more important than in urban slums, where humans are typically crowded, often inherently vulnerable, and live in close proximity with animal and environmental reservoirs of infectious agents. Leptospirosis, a spirochetal zoonosis, has emerged as an important health problem in tropical slum settlements. The disease is associated with life-threatening clinical manifestations for which case fatality is >10%. Effective preventive measures are urgently needed, as the population residing in slums will double from one to two billion in the next 20 years. In urban slums of Brazil and many other temperate and tropical cities, leptospires are transmitted to humans through contact with environments contaminated by infectious urine shed from rodents. Norway rats (Rattus norvegicus) are the principal reservoir host for leptospirosis and the prevalence of chronic infection of their kidneys exceeds 50% within rat populations of Salvador Brazil. However, our epidemiologic understanding of urban leptospirosis is limited and knowledge of how rat abundance contributes to risk of human infection are critical to inform expensive rat-control efforts, which in Brazil and other developing have been uniformly ineffective. The proposed study aims to~ 1) Assess, quantify and improve upon the three currently used methods [trapping counts, tracking board and rodent infestation surveys] for estimating R. norvegicus abundance in urban slum conditions and 2) Validate these metrics as a predictive measure of leptospirosis infection and distribution amongst a longitudinally-studied cohort of urban slum dwellers. To succeed, novel statistical and geostatistical methods will be developed to generate integrated models of the best combination of rat abundance measures with geocoded data (demographics, socioeconomic status and risk behaviors obtained from the cohort study coupled with environmental data and qualitative rat-survey information) to produce predictive GIS-generated map surfaces. The results of the proposed study will elucidate the dynamics between rat abundance and risk of human leptospirosis infection/disease. The results will provide evidence-based results to better inform rodent control campaigns. The methods developed in the proposed study will be applicable for a wide range of both epidemiological and ecological studies on zoonotic disease. Furthermore the proposed study will provide the candidate with training in designing and implementing field investigations in resource poor settings, applying advanced spatio-temporal statistical methodologies to understand disease transmission, and developing a multidisciplinary skill set and collaborations which will enable her to pursue her long-term goal of addressing zoonotic diseases through research in eco-epidemiology.